KMA Weather

You set up a dedicated MCP server that connects the Korea Meteorological Administration’s short-term forecast Open API. It exposes current weather data and short-range forecasts through simple, URL-based resources your MCP client can query, making it easy to integrate weather information into your applications.

How to use

You interact with this MCP server through standard MCP endpoints. Retrieve current weather data for Seoul by requesting the resource weather://seoul/now. For other locations, specify latitude and longitude in the URL as weather://{latitude}/{longitude}/now. You can also obtain short-range forecasts using the provided tools that the server exposes through the MCP framework.

The available data sources include current conditions and short-term forecasts. When you query these resources through your MCP client, you receive structured weather information suitable for dashboards, alerts, or decision-making in your applications.

How to install

Prerequisites you need installed on your system include Node.js and Python, since the MCP client tooling and the server runtime rely on these environments.

Prepare the environment file with your API key decoded value. Create or update the file at the project root with:

KMA_API_KEY_DECODED=your_decoding_key_here

If you are configuring the Claude Desktop client to run this MCP server, add a configuration entry that points to the server script and provides the required environment variable. Use the following settings as a template, replacing the absolute path with the actual location of your server script.

{
  "mcpServers": {
    "kma-weather": {
      "command": "uv",
      "args": [
        "run",
        "-q",
        "--with",
        "mcp[cli]",
        "--with",
        "httpx",
        "--with",
        "fastmcp",
        "--with",
        "python-dotenv",
        "/ABSOLUTE/PATH/TO/Public API to MCP Converter Agent/kma-weather-mcp/src/server.py"
      ],
      "env": {
        "KMA_API_KEY_DECODED": "YOUR_KEY_HERE (Optional if using .env file in project dir)"
      }
    }
  }
}

Start the server using the configured command. The runtime expects the same arguments shown here, so do not omit any part of the command. This example uses the path placeholder shown in the configuration snippet.

uv run -q --with mcp[cli] --with httpx --with fastmcp --with python-dotenv /ABSOLUTE/PATH/TO/Public API to MCP Converter Agent/kma-weather-mcp/src/server.py

Additional notes

Environment variables shown in the configuration must be available to the server at runtime. If you place them in a project-level .env file, the server will automatically load them through python-dotenv.

Development workflows can also run via a development inspector, as shown in the example: npx @modelcontextprotocol/inspector uv run src/server.py.

Available tools

get_ultra_short_term_forecast

Fetches the ultra-short-term forecast for the next 6 hours, including precipitation probability and sky conditions.

get_village_forecast

Retrieves short-term forecasts from today through the day after tomorrow.